The framework model of the folding transition state explains an intriguing observation: folding rates of kinetically two-state proteins are correlated with contact order, a measure of the extent of local versus non local interactions in the native protein. For framework folding all interactions found in the transition state ensemble are native-like, so that native structure can guide design of mutant proteins with perturbed transition states. The framework model assumes that early folding events are encoded locally in secondary structure preferences. Secondary structure forms rapidly - on a nanosecond time scale. These locally-formed marginally stable structural elements are stabilized further in the later stages of folding by tertiary interactions formed on a microsecond timescale via a diffusion-collision process. Native protein is formed on a relatively slow timescale of hundreds of microseconds to several milliseconds as the final elements of structure are locked in and water is excluded from the hydrophobic core. Although specific aspects of the framework model have been tested in specific proteins and the observed trends for topologically distinct proteins are highly suggestive, simultaneous tests of all contributing factors within the same protein have never been carried out. I propose tests of all four aspects of the framework model in the same protein, iso-2 cytochrome c. All the tests involve mutagenic perturbations of transition state stability that have predictable effects on folding rates. Aspects of the transition state ensemble to be investigated include: 1) intrinsic stability of native-like substructures; 2) stability of tertiary contacts between substructures; 3) loop length dependence of chain entropy losses on closing polypeptide loops; 4) sequence preferences for turn and loop formation. The effects of mutational perturbations on folding rates will provide support for framework assembly; or suggest alternative models for transition state stabilization. The tests will use engineered variants of (yeast) iso-2 cytochrome c lacking kinetic blocks to folding. Data analysis will make use of the phi-value formalism, and Kramers' theory of reaction rates in solution. To span the microsecond to tens of milliseconds timescale of folding, rates will be measured by (microsecond) temperature jumps. Scanning calorimetry will be used to measure global stability, and native-state hydrogen exchange to quantify changes in local stability of perturbed substructures.